COMMENTARY
Ecological Complexity and West Nile Virus Perspectives on Improving Public Health Response Daniel G.C. Rainham, MSc
ABSTRACT The emergence of West Nile Virus, as well as other emerging diseases, is linked to complex ecosystem processes such as climate change and constitutes an important threat to population health. Traditional public health intervention activities related to vector surveillance and control tend to be reactive and limited in their ability to deal with multiple epidemics and in their consideration of population health determinants. This paper reviews the current status of West Nile Virus in Canada and describes how complex systems and geographical perspectives help to acknowledge the influence of ecosystem processes on population health. It also provides examples of how these perspectives can be integrated into population-based intervention strategies. MeSH terms: Disease transmission; ecological systems; geographic information systems; systems theory; primary prevention
RÉSUMÉ L’apparition du virus du Nil occidental, à l’instar d'autres maladies émergentes, résulte de processus écosystémiques complexes, comme le changement climatique, et représente une importante menace pour la santé. Les interventions traditionnelles de santé publique, telles que la surveillance et le contrôle des vecteurs, demeurent limitées dans leur capacité d’endiguer les épidémies ou d’agir sur leurs déterminants. Cet article examine la situation canadienne et l’état actuel des connaissances sur le virus du Nil occidental. En s’appuyant sur une analyse des systèmes complexes et une analyse spatiale, il décrit comment les écosystèmes influencent la santé des populations et, en conséquence, quelles stratégies d’intervention populationnelle pourraient être envisagées.
Correspondence: Daniel Rainham, McLaughlin Centre for Population Health Risk Assessment, Institute of Population Health, University of Ottawa, One Stewart Street, Ottawa, ON K1N 6N5, Tel: 613-562-5800 ext. 2092, Fax: 613-562-5380, E-mail:
[email protected] Acknowledgements: The author gratefully acknowledges support from the Canadian Institutes of Health Research in the form of a Doctoral Research Award and from the McLaughlin Centre for Population Health Risk Assessment. The paper has also benefited from the helpful comments of two anonymous reviewers. JANUARY – FEBRUARY 2005
est Nile Virus (WNV) as well as other emerging infectious diseases have been characterized as important threats to population health,1-3 and constitute a major focus of population health intervention and policy research.4 Fluctuations in climatic variables – most notably temperature, precipitation, humidity and wind – influence the distribution and processes among biological organisms linked to the spatial and dynamic variability of infectious diseases.5 Disease surveillance and vector control activities must acknowledge the inherent complexity characteristic to ecosystem-based transmission cycles, as well as the complex interactions among the determinants of population health. 6-8 Population-based intervention strategies will require unconventional analytic tools and alternative conceptual foundations if they are to be successful at tackling public health risks characterized by multiple interactions, non-linear rates of change, and spatio-temporal influences. This paper discusses how complex systems and geographical perspectives can contribute to the design of population health interventions for West Nile Virus, and by extension, other emerging infectious diseases.
W
Transmission and expansion of West Nile Virus The sudden appearance of severe and fatal encephalitis in dying corvids (i.e., crows, ravens, blue jays) and in humans in New York City in 1999, and in Canada in 2001, provides a compelling example of the expanding range of emerging diseases and a corresponding expansion in diagnostic capacity to detect emerging health threats. West Nile Virus transmission cycles and maintenance mechanisms are usually very complex, and the virus can easily become established if efficient vectors, suitable amplifying hosts, and reliable overwintering mechanisms are available.9 Indigenous to Africa, Asia, Europe, and Australia, and now North America, the origin of the original strain of WNV is the Middle East, but the mode of introduction is unknown. 10 Humans are incidental (dead end) hosts since they can become ill as a result of the virus, but do not develop viremia that is required for the continuation of the transmission cycle if bitten by another mosquito.11 West Nile Virus constitutes a serious threat to human population health and CANADIAN JOURNAL OF PUBLIC HEALTH
37
ECOLOGICAL COMPLEXITY AND WEST NILE VIRUS
well-being. Annual human case incidence now ranks WNV second only to LaCrosse encephalitis virus as the leading cause of reported human arboviral encephalitis in the United States.11,12 The 2002 WNV epidemic in the United States was the largest arboviral epidemic documented in the western hemisphere, with the greatest intensity in the central states and Great Lakes region amounting to a total of 3,389 human cases across 2,289 counties in 44 states.13 In 2002, Health Canada documented WNV activity in five provinces (Manitoba, Nova Scotia, Ontario, Quebec, and Saskatchewan) with 343 cases reported for Ontario.14 However, doctors with the Canadian Infectious Disease Society suspect nearly 1,000 symptomatic human cases of WNV in Ontario in 2002, much larger than the official reports.15 West Nile Virus may also be transmitted indirectly by blood and blood product transfusion and organ transplantation,16 directly via parenteral transmission from breastfeeding,17 as well as occupationally among medical laboratory workers through percutaneous injection.18 A complex systems perspective A complex systems perspective recognizes that fundamental ecosystem (physical, biological and social) processes affect the survival and reproduction of all living organisms.19 A long history of coevolution has governed the relationship among vectors, microorganisms (viruses for example), their hosts, and the environmental conditions within which they interact.20 Even with such history, the diversity, distribution, and effects of infectious diseases are in constant flux. For example, ecosystem processes such as interactions between vectors and hosts are inherently complex since each element may adapt and/or reorganize in response to intervention. Enhancing infectious disease surveillance and integrated vector control will require acknowledgement of the complexity underlying ecologically-linked transmission cycles.21 Complexity in any ecosystem arises from inter- and intra-specific interactions among individuals, interactions across levels of organization, and the interaction of organisms with the environment over space and time.22 Complex systems theory incorporates notions of adaptation 23 and emergence, 24 and has been adopted by the 38
REVUE CANADIENNE DE SANTÉ PUBLIQUE
TABLE I Definitions* of Key Terms Term Complex Systems
Attractors
Dissipative Structure Leverage points† Emergence
Definition There is no one accepted definition of complex systems. However, the properties of a complex system are: openness (to material and energy flows), nonequilibrium (exist in states away from equilibrium), non-linear (behave as a whole and not amenable to reductionist approaches), selforganizing (non-Newtonian, characterized by feedback loops and surprise), hierarchical (systems within systems at different scales), catastrophic behaviour (normal but could involve thresholds), chaos (prediction always limited no matter how powerful our models or computers). In short it means irreducible uncertainty. Dynamic complex systems are inherently chaotic and unstable, but they usually settle down into one of a number of possible steady states. These steady states are called attractors. Steady systems are usually found near attractors due to feedback loops. A system that becomes more differentiated requires more energy to sustain it than the simpler system it replaces, and becomes less efficient in the use of energy. There are levers, or places within a complex system (such as a firm, a city, an economy, a living being, an ecosystem), where a small shift in one thing can produce big changes in everything. When a system is able to change profoundly and generates new properties and structures from existing ones, it is said to have emergent properties. The new system is qualitatively different and did not result from traditional Newtonian mechanics, but rather from simple non-linear interactions.
* An excellent introduction to complex adaptive systems and ecosystem approaches can be found on the personal website of Dr. James Kay at www.jameskay.ca † Look for the Winter 1997 essay entitled “Places to Intervene in a System” by Donella Meadows at www.wholeearthmag.com
Centers for Disease Control’s Syndemic Prevention Network to examine the synergisms and interactions evident in the connections between higher-order, healthinfluencing phenomena.25 Complex systems consist of dissipative structures and attractors. Dissipative structures and processes describe the state of any system from a position far from equilibrium to one that is close to equilibrium. 26 Tornados or autocatalytic chemical reactions are good examples. Attractors are specific points or features within a system that exert leverage on future evolution.27,28 (See Table I for definitions of specialized terms used in this paper.) Figure 1 represents graphically an application of complex systems thinking adapted to understanding the emergence of WNV in relation to human activities. The model proposes four states: exploitation and rapid growth, conservation, release (collapse), and reorganization, as well as a list of potential attractors likely to be found at each state. By following the figure eight pathway in the direction of the arrows, it is possible to trace the relationships between human-focussed development activities and resulting reductions in system resilience leading to the emergence of infectious disease and possible public health responses. Emergent properties may also be positive by promoting dynamic interaction between different attractors
within the system. For example, the creation of intersectoral and crossjurisdictional linkages among traditionally disparate groups, such as provincial health authorities and federal wildlife research agencies, reveals efforts to understand the dynamic nature of human-ecosystem interactions that foster the conditions necessary for opportunistic viruses.29 Typical populationbased interventions are adaptive and generally located in the pathway between the system states of release and reorganization, a position where the system is unstable and behaving chaotically. Intervention in complex environments should take place at multiple points along the pathway between system states, even though intervention targets may be difficult to identify. Adopting a systems perspective is useful for identifying those attractors and structures more or less amenable to change, but does not guarantee successful intervention. Geographic perspectives on infectious disease and intervention Patterns of differentiation in complex systems have a spatial dimension that is characterized by patterns of territoriality and spatial heterogeneity. Incorporating geographical perspectives into populationbased intervention strategies forces researchers to contemplate location and place, as well as the relationship between disease and population health patterns VOLUME 96, NO. 1
ECOLOGICAL COMPLEXITY AND WEST NILE VIRUS Reorganization - small multiple inputs - capital dependent - system connections not always evident - system emergence (new)
Conservation - human culture & dominance - social structures - market focus vs. ecology - less diversity - reduced resilience
To another system form
Exploitation - population growth - urban development - deforestation - resource dependence
Figure 1.
Release - disorganization - emerging diseases (WNV) - social disturbances - system stress
The emergence of West Nile Virus from a complex systems perspective. Follow the figure eight pathway in the direction of the arrows to trace the relationships between human-focussed development activities and resulting reductions in system resilience leading to the emergence of infectious disease. Intervention activities should take place at multiple points along the path and not only after the release state (point where WNV becomes a public health threat). Figure is a modification of Hollings’ Four Box Model.
across scales and social groups. For example, the social, cultural and physical environment exacts a “nested” effect on the health of populations: features of the micro-, meso- and macro-environmental scale interact to create health in context.30 Understanding the spatial dynamics of West Nile Virus is critical for the determination of its emergence in new locations or among potential vectors and hosts. Often the emergence of infectious disease is characterized by complexity and chaotic processes. For example, the leading edge of many infectious disease epidemics is often jagged and irregularly shaped, reflecting spatial variation in the distribution of susceptible hosts, variation in localized contact rates, long distance vector migration patterns, or variations in intervention intensity and duration (e.g., intervention activities tend to vary considerably among health units).21 Analyses of pathogen dispersal characteristics display distinct fractal patterning inherent to non-linear processes.31 Fractal models have been proposed to explore the spatial shape of social networks in the process of HIV transmission.32 Effective health intervention approaches tend to be collaborative and multi-level in nature.33,34 Implementation of multi-level interventions increases the amount of complexity creating challenges for finding established measures for variables across JANUARY – FEBRUARY 2005
scales and for intervention evaluations. These challenges can be overcome by employing multi-level models using geographical aggregations of space, 35 or by assessment of patterning of emerging disease at distinct levels of aggregation. 36 Spatial patterning assessments have been used to analyze the spatial distribution of mosquito bites on the human body.37 Such information would be useful for tailoring intervention strategies to reduce personal exposures. Spatial approaches have also been used to assess the relationship between intervention coverage and public health unit areas. It is highly unlikely that spatial patterning of disease vectors or locations of transmission hotspot areas will fit nicely into existing public health unit boundaries that are artificially determined from political/administrative processes or data collection regimes.38 Adoption of a spatialized ontology into the design of intervention for West Nile Virus may well enable the development of surveillance and control strategies based on meaningful biological or ecologically relevant units across multiple scales and ecosystem processes. The concept of location is critical to understanding the intersection of population health determinants and ecologicallyrelevant intervention activities.39 Mostly absent from the population health discourse is discussion about the role of geo-
matics as a common conceptual framework for studying the location of health aetiology and intervention. Geomatics comprises a multi-disciplinary approach to the science and technology associated with gathering, storing, analyzing, interpreting, modeling and distribution of georeferenced information.40 A major strength of using geomatics for population-based intervention is the ability to collect and merge large amounts of varied data into a single database. For example, by using a geomatics framework, it would be possible to overlay and integrate multiple data sets from various scales to describe health status, examine interactions among health determinants, and help target intervention strategies. Such data may include, but are not limited to: population data (e.g., socioeconomic, census information), environmental and ecological data (e.g., pollution, vegetation, remotely sensed information), topography, hydrology and climate data (particularly useful for understanding vector-borne disease), land-use and public infrastructure information (e.g., schools, drinking water supplies), transportation networks, health infrastructure and epidemiologic data, and most any other data.41 Successful interventions oriented towards the reduction of risk from malaria in Africa have adopted a geomatics approach to assess the type and severity of malaria transmission, and to tailor appropriate control measures according to the capacity and needs of specific regions.42 Although many public health ministries and units in Canada use geographic methods for infectious disease control, it is not clear whether these methods extend much beyond visual display of the data or if geographic thinking extends to the actual design of health interventions. CONCLUSIONS West Nile Virus is now endemic in Canada and constitutes a serious concern for population health and well-being. Although the number of confirmed vectorborne cases is small, many of these cases go unreported. Transmission cycles are connected to local and global climate variability and complex ecosystem dynamics. A complex systems approach is useful for thinking about population-based intervention to WNV because: 1) it acknowledges CANADIAN JOURNAL OF PUBLIC HEALTH
39
ECOLOGICAL COMPLEXITY AND WEST NILE VIRUS
relationships between ecosystem-level interactions and the emergence of infectious disease, and 2) it helps to identify structures and processes that are more or less agreeable to change at multiple points and scales within a system. Geographical perspectives also have great utility in the design of intervention strategies. Gaining insight into the spatial processes influencing the spread of West Nile Virus will ultimately aid in the design of more efficient surveillance and monitoring activities. Many vector-borne and other emerging infectious diseases exhibit a spatial patterning or clustering which may vary according to the scale under investigation. Interventions can use location-specific information to modify intervention planning by maximizing resources for regions with severe outbreaks, and to incorporate traditionally disparate information and data relevant to infectious disease control activities. Ideally, public health workers, physicians, and other health care providers might benefit from the adoption of these perspectives into infectious disease intervention planning and may be made more aware of the relationships among ecosystemlevel processes, human activities, and the population health risks associated with West Nile Virus. REFERENCES 1. Lederberg J, Shope RE, Oaks SC (Eds.) Emerging Infections: Microbial Threats to Health in the United States. Washington, DC: National Academy Press, 1992. 2. Centers for Disease Control and Prevention (CDC). Preventing Emerging Infectious Diseases: A Strategy for the Twenty-first Century. Atlanta, GA: CDC, 1998. 3. Charron DF. Potential impacts of global warming and climate change on the epidemiology of zoonotic diseases in Canada. Can J Public Health 2002;93(5):334-35. 4. Institute of Medicine (IOM). The Resistance Phenomenon in Microbes and Infectious Disease Vectors: Implications for Human Health and Strategies for Containment. In: Knobler SL, Lemon SM, Najafi M, Burroughs T (Eds.), Forum on Emerging Infections, Board on Global Health. Washington, DC: National Academy Press, 2003. 5. Elias SA. Insects and climate change: Fossil evidence from the Rocky Mountains. Bioscience 1995;41(8):552-59.
40
REVUE CANADIENNE DE SANTÉ PUBLIQUE
6. Morse SS. Emerging viruses: Defining the rules for viral traffic. Perspect Biol Med 1991;34(3):387-409. 7. Levins R, Awerbuch T, Brinkmann U. The emergence of new diseases. Am Sci 1994;82:59-60. 8. Epstein PR. Emerging diseases and ecosystem instability: New threats to public health. Am J Public Health 1995;85(2):168-72. 9. Calisher CH. West Nile Virus in the new world: Appearance, persistence, and adaptation to new econiche – an opportunity taken. Viral Immunol 2000;13:411-14. 10. Campbell GL, Marfin AA, Lanciotti RS, Gubler DJ. West Nile Virus. Lancet: Infect Dis 2002;2:519-29. 11. Craven RB, Roehrig JT. West Nile Virus. JAMA 2001;286:651-53. 12. Centers for Disease Control. Epidemic/Epizootic West Nile Virus in the United States: Revised Guidelines for Surveillance, Prevention, and Control. Workshop on health in Charlotte, North Carolina. Atlanta, GA: CDC, 2001. 13. Nash D, Mostashari F, Fine A, Miller J, O’Leary D, Murray K, et al. The outbreak of West Nile Virus infection in the New York City area in 1999. N Engl J Med 2001;344(24):1807-14. 14. Ministry of Health and Long-Term Care (MOHTLC). West Nile Virus Preparedness and Prevention Plan for Ontario, 2003. Toronto, ON: MOHTLC, 2003. 15. Environmental Risk Analysis Program. What’s going on with West Nile Virus? Cornell University, Center for the Environment. September 17, 2003. Available at http://environmentalrisk.cornell.edu/WNV/. Accessed February 2004. 16. Centers for Disease Control. Update: Investigations of West Nile Virus infections in recipients of organ transplantation and blood transfusion. MMWR 2002;51(39):879. 17. Centers for Disease Control. Possible West Nile Virus transmission to an infant through breast feeding. MMWR 2002;51(39):877-78. 18. Centers for Disease Control. Provisional surveillance summary of the West Nile Virus epidemic – United States, January-November 2002. MMWR 2002;51:1129-33. 19. Levin SA. Ecosystems and the biosphere as complex adaptive systems. Ecosystems 1998;1:431-36. 20. Ewald PW. Evolution of Infectious Disease. New York, NY: Oxford University Press, 1994. 21. Wilson ML. Ecology and infectious disease. In: Aron JL, Patz J (Eds), Ecosystem Change and Public Health: A Global Perspective. Baltimore, MD: Johns Hopkins University Press, 2001;283324. 22. Hartvigsen G, Kinzing A, Peterson G. Use and analysis of complex adaptive systems in ecosystem science: Overview of special section. Ecosystems 1998;1:427-30. 23. Jansenn M. Use of complex adaptive systems for modeling global change. Ecosystems 1998;1:45763. 24. Levin SA. Concepts of scale at the local level. In: Logan JR, Field CB (Eds.), Scaling Physiologic Process: Leaf to Globe. New York, NY: Academic Press, 1993;7-19. 25. Centers for Disease Control. Introduction to the Syndemics Prevention Network. Atlanta, GA: CDC. July 16, 2003. Available at
26.
27. 28. 29. 30.
31. 32.
33. 34. 35. 36. 37. 38.
39.
40. 41.
42.
http://www.cdc.gov/syndemics. Accessed on January 24, 2004. Higginbotham N, Albrecht G, Connor L. Health Social Science: A Transdisciplinary and Complexity Perspective. Melbourne, Australia: Oxford University Press, 2001. Prigogine I, Stengers I. Order Out of Chaos: Man’s New Dialogue with Nature. New York, NY: Bantam Books, 1984. Greenhalgh T. Change and complexity. Br J Gen Pract 2000;50:514-15. Peterson G. Political ecology and ecological resilience: An integration of human and ecological dynamics. Ecol Econ 2000;35:323-36. Andrews G. The ecology of risk and the geography of intervention: From research to practice for the health and well-being of urban children. Ann Assoc Am Geogr 1985;75(3):370. Kenkel NC, Irwin AJ. Fractal analysis of dispersal. Abstr Bot 1994;18:79-84. Wallace R. A fractal model of HIV transmission on complex sociogeographic networks: Towards analysis of large data sets. Environ Plan A 1993;25:137-48. Green LW, Richard L, Potvin L. Ecological foundations of health promotion. Am J Health Prom 1996;10:270-81. Grzywacz J, Fuqua J. The social ecology of health: Leverage points and linkages. Behav Med 2000;26:101-15. Duncan C, Jones K, Moon G. Health-related behaviour in context: A multilevel modelling approach. Soc Sci Med 1996;42(6):817-30. Allen CR, Forys EA, Holling CS. Body mass patterns predict invasions and extinctions in transforming landscapes. Ecosystems 1999;2:114-21. Enserink M. What mosquitos want: Secrets of host attraction. Science 2002;298:90-92. Patil GP, Bishop J, Myers WL, Taillie C, Vraney R, Wardrop DH. Detection and delineation of critical areas using echelons and spatial scan statistics with synoptic cellular data. Available at http://www.stat.psu.edu/~gpp/PDFfiles/TR2002 -0501.pdf. Accessed on April 3, 2003. Sallis JF, Owen N. Ecological models of health behaviour. In: Glanz K, Rimer BK, Lewis FM (Eds.), Health Behavior and Health Education: Theory, Research, and Practice. San Francisco, CA: Jossey-Bass, 2002;462-84. Geomatics Canada. Available at http://www.nrcan.gc.ca/geocan/index_e.html. Accessed on March 21, 2003. Gatrell AC, Dunn CE. Geographical information systems and spatial epidemiology: Modelling the possible association between cancer of the larynx and incineration in North-West England. In: De Lepper MJC, Scholten HJ, Stern RM (Eds.), The Added Value of Geographic Information Systems in Public and Environmental Health. Boston, MA: Kluwer Academic Publishers, 1995;215-33. MARA/ARMA (Mapping Malaria Risk in Africa). August 12, 2003. Available at http://www.mara.org.za. Accessed on December 13, 2003.
Received: March 15, 2004 Accepted: July 30, 2004
VOLUME 96, NO. 1
